IEEE Talks IoT: Joe Weinman
Joe Weinman is the author of Cloudonomics: The Business Value of Cloud Computing, the forthcoming book Digital Disciplines, and is the chair of the IEEE Intercloud Testbed initiative.
Question: How do you define the Internet of Things?
Weinman: AI worked for a network service provider for decades, so I can’t say that I love the term since there are so many network technologies in addition to “the Internet.” Still, the notion of things that are connected to identical things, other things, cloud computing, storage, and applications, and people is a powerful one that represents the latest wave in information technology. For example, wind turbines can connect to other wind turbines, but also the electric power utility grid and optimizing software. To me, then, the “Internet of Things” is all about intelligent, adaptive, digitalized sensors, actuators, devices and systems that connect across a variety of networks to other devices and systems and to back-end services, including computing, partner ecosystems, and social networks.
Question: What do you think are the challenges and opportunities?
Weinman: I think the opportunities are limitless. I believe that there are four main “digital disciplines”—where information technologies meet generic business strategies. They are “information excellence,” where information is leveraged to optimize processes and asset utilization; “solution leadership,” where products are integral components of connected product-service systems focused on business outcomes; “collective intimacy,” where data from all customers is used to create targeted recommendations for each customer, whether they be movie recommendations or patient-specific therapies; and “accelerated innovation,” where monolithic closed industrial laboratories are supplanted or complemented by open innovation, crowdsourcing, idea markets, and contests and challenges.
Connected things play a key role in each of the digital disciplines. For example, information excellence— whether for port logistics or flexible manufacturing or wind turbine farms—requires intelligent devices, which GE refers to as “Brilliant Machines” which feed data in real time to optimizing algorithms and heuristics. Solution leadership—such as connected cars, connected jet engines, or athletic shoes with built-in pressure sensors—begins with things that then connect to back-end services such as navigation or concierge services for cars; flight optimization services for engines; or activity tracking and coaching for sneakers. Collective intimacy requires some sort of device to collect data for further analysis via collaborative filtering algorithms, for example. Finally, innovation can often require big data, created by machines. What the late Jim Gray of Microsoft called “the fourth paradigm of science” involves looking for correlations in large data sets, e.g., the huge amount of data created by CERN’s Large Hadron Collider.
Question: What are the technological innovations and advances and why are they important?
Weinman: There are many key innovations that have already occurred and more that will. Low cost, low power processing, storage, and networking are examples. Device power is one example: processing elements need to become more energy efficient, and battery life needs to be either extended or complemented through “energy scavenging,” e.g., generating power through piezoelectricity based on microscopic vibrations in the environment. In networking, efficient low-latency protocols are needed, as well as architectures such as mesh topologies to enable more ubiquitous networking. Nanotechnology is important, as well as entirely new areas of technology, such as the memristor, which promise new efficiencies in price/power/performance for IoT infrastructures.
Question: What will be the process to get the IoT or certain technologies deployed or implemented?
Weinman: I don’t think it will vary very much from any other technology. On the one hand, you have very practical aspects with immediate payoffs that companies will implement under a standard business case process. For example, a technology that enables process optimization will have a typical business case where a front-end capital investment in devices and capitalized software is more than made up for by a longer-term stream of savings. At the other extreme, you’ll have plenty of start-ups with innovating products, services, and business models, which burn through capital and generate a lot of failures, like Pets.com, but a few who create unimaginable wealth and cause us to rethink fundamentals. There is virtually infinite opportunity for reinvention of simple things. The Nest thermostat shows how elegance in design can intelligence, adaptability, and connectivity can create billion dollar businesses virtually overnight. Not everything has to have a financial payback associated with it: the pet rock was very successful in its day and there are plenty of laughing baby videos on YouTube.
So, some mix of “boring” pragmatic implementations and creative applications that capture the imagination of the mass market will be what drives adoption.
Question: What are the anticipated costs/risks and benefits/impacts?
Weinman: The most interesting benefits to me are how the Internet of Things can enable creative business strategies and competitive differentiation. I think the main risks are security and unintended consequences. Security and privacy are always issues in information technology, and they’re issues that will never go away. If today we are concerned about an intruder accessing a private love letter or text or Facebook photo, think about the future when the intruder will access a security video feed from our home. Or take over our car while it is hurtling down the highway. Or, worse yet, disable or otherwise maliciously impact critical elements of national infrastructure.
Unintended consequences are another issue. We’ve seen cases such as the “flash crash” where systems interacting with each other have unusual system dynamics leading to sudden chaotic behavior. How can one anticipate all the possibilities when different products and devices from different vendors interact with each other in complex ways?
Another issue is standards. No one wants a world of billions of devices, none of which can talk to each other. We want an Internet of Things, not a random collection of isolated Islands of Technology. Having vendor-agnostic standards to enable secure interoperability, prevent unintended system dynamics, and achieve substantial benefits for consumers and businesses is an important goal.